The second round of the French election, to be held on July 7, carries some interesting lessons about democracy. In each circumscription where no candidate obtained more than 50% of the votes in the first round, those who got more than 12.5% are allowed to run in the second round. A political party or coalition whose candidate came third or lower may have an interest (and an informal obligation under electoral agreements) to pressure him to drop out in order not to split the votes among the two major candidates in case the election of one of them would be prejudicial to its post-election position in the National Assembly. “Centrist” parties allied with the leftist New Popular Front to try to block the “far-right” National Rally. (I put “far-right” in scare quotes because NR is not unquestionably farther to the right than NPF is to the left, and many of their statist proposals are similar.) This strategy led 224 candidates to drop out in the 577 circumscriptions. (See “French Elections: 224 Candidates Have Officially Withdrawn from the Second Round,” Le Monde, July 2, 2024.)

The purpose of a second round is to increase the chances (or to guarantee, depending on the exact setup) that the elected candidate will be able to claim to represent the “will of the people,” that is, 50%+1 of the individuals making up “the people.” One might think that, for a worshipper of democracy, removing one option from the voters’ menu would be sinful. Technically, it violates the condition called “neutrality” in democratic theory, for it favors some options over others. In reality, though, limiting options presented to the voters necessarily happens all the time, one way or another, if only because there are zillions of possible collective (political) choices; each voter potentially has his own ideal option.

For any single voter, voting choice limitations are inconsequential because his vote, whatever the menu, is not decisive. He (including she, of course) would stay home and the winner would not change. However, a political strategy of making one candidate drop out may change the collective choice resulting from the election, compared to what it would otherwise have been. The contradictions and inconsistencies of democratic mythology are numerous.

No democratic gadgetry can make an election or referendum better express “the will of the people,” which does not exist anyway. As I noted in a previous post, different democratic voting methods can achieve widely different results. Interpreting the work of Donald Saari (“Millions of Election Outcomes from a Single Profile,” Social Choice and Welfare, 1992), Gordon Tullock wrote (in Government Failure: A Primer in Public Choice, 2002):

Many different voting rules are used in the world and each leads to a somewhat different outcome. Saari has produced a rigorous mathematical proof that for a given set of voters with unchanged preferences, any outcome can be obtained with at least one voting method.

Combining all that with the Condorcet Paradox and its contemporary extensions, it would be an error to search for the unfindable majority. A majority is only one possible majority among many, depending on the voting system and back-office politics, not to mention the frequent bureaucratic influence on the political agenda. As political scientist William Riker would put it, democratic decisions are either dictatorial or “arbitrary nonsense, at least some of the time” (see his Liberalism Against Populism, 1982).

The non-negligible benefit of constitutional democracy (“constitutional” means “limited”) is to offer voters, when enough are dissatisfied with their rulers, a low-cost means to get rid of them. Liberal democracy (which, in its classical sense, means constitutional democracy), Riker writes, allows for “an intermittent and sometimes random popular veto” that has some capability of restraining “official tyranny.” We must not ask too much from democracy.

As much as the limitation of the options presented to an electorate is unavoidable, the constant limitation of individual choices by collective choices is not the only imaginable state of the world. It is generally inefficient or immoral or both. A collective choice removes many options from the opportunity sets of individuals. It has a direct effect on the choices of all individuals who would have done what is now forbidden. This, not democratic mythology or gadgetry, is the important issue.


I instructed ChatGPT to “generate an image illustrating democracy.” I did not tell “him” anything else. He described his image (the featured image of this post, reproduced below) as follows: “A vibrant and diverse group of people standing together in a large open space, each holding a different flag representing various countries around the world. In the center, there is a large, ornate ballot box on a raised platform, symbolizing democracy. Above the scene, a bright sun shines, casting a hopeful and unifying light over the crowd. The background includes iconic global landmarks such as the Eiffel Tower, Statue of Liberty, and the Great Wall of China, representing international unity and cooperation.” It is a vacuous concept of democracy: democracy is nice and good; but it is probably widely shared, as the bot’s database attests. (“He” produced a second image, at the same level of vacuity.)

Democracy according to DALL-E

Democracy is nice, good, and vacuous, if we believe DALL-E